ECG Signal Analysis Using Wavelet Transform

ECG signal plays an important role in the primary diagnosis and analysis of heart diseases. The feature of ECG signal with time-varying morphological characteristics needs to be extracted by signal processing method because there are not clearly visible in the graphical ECG signal. For analyzing this kind of signal wavelet transforms are a powerful tool. In this thesis paper, an algorithm for automatic ECG signal feature extraction was evaluated. For feature extraction multi-resolution wavelet transform is used. Text formatted ECG signals are taken from the MIT-BIH arrhythmia database. For wavelet transform, daubechies wavelets were used because the scaling functions of this wavelet filter are similar to the shape of the ECG. In the first step, the ECG signal was denoised by removing the corresponding higher scale wavelet coefficients. Then the R wave peaks were detected which have higher dominated amplitude. These detected R peaks were later used to detect the other peak as P, Q, S, T and also the zero crossing level. From the different peaks, the features of the ECG signal were extracted. Depending on different features, different types of abnormality are classified. The ECG is nothing but the recording of the heart's electrical activity that is generated by depolarization and repolarization of the atria and ventricales(01). ECG is an important tool for the primary diagnosis of heart disease; it shows the electrophysiology of the heart and the ischemic changes that may occur like the myocardial infection, conduction defects and arrhythmia (02). One cardiac cycle in an ECG signal consists of the P-QRS-T waves . Most of the clinically useful information in the ECG is found in the intervals and amplitudes defined by its features. The ECG feature extraction system provides fundamental features (amplitudes and intervals) to be used in subsequent automatic analysis. Algorithms for ECG feature extraction are difficult to produce due to temporal variations from physiological conditions and the presence of noise. Beat or QRS complex detection is the most important part of an ECG feature extraction system. R wave peak is the QRS complex designator, so peak detection algorithms are required.